Zoom in screen with simultaneous prompting to teach young children with autism

Abstract

Telehealth is a broad term that describes the use of communication technologies to deliver diagnoses and educational services. Recent guidelines for practicing telehealth indicate that interventions directly delivered to individuals with autism spectrum disorder (ASD) via telehealth may be appropriate and effective. However, as few studies have directly delivered interventions to individuals diagnosed with a disability via telehealth, more research is needed. This study examined the effects of telehealth-delivered simultaneous prompting combined with observational learning in young children with ASD. Two children aged 5 years who were diagnosed with ASD participated in this study. The students and implementers were located in different physical settings. A multiple-probe design across behaviors replicated across participants was used to examine the effects of the procedure. The students acquired the target and observational responses. Importantly, both students maintained the acquired responses one, two, and six weeks after the intervention and generalized them across different conditions. The students’ teachers had positive opinions regarding the procedures and outcomes. The findings of this single-subject study have important implications for delivering interventions directly to young children with ASD via telehealth. The study’s limitations and implications for future research are also discussed.

Introduction

Telehealth is defined as the use of communication technologies to deliver diagnoses or educational services (Ferguson et al. 2020). Telehealth service delivery has become increasingly popular for teaching individuals with autism spectrum disorder (ASD) owing to the COVID-19 pandemic (Pellegrino and Reed 2020). Telehealth is used in the provision of applied behavior analytic services for this population (Lerman et al. 2020). Applied behavior analysis (ABA) is an applied science aimed at evaluating environmental variables and their relationships to target behaviors and design interventions accordingly. Practices derived from the principles of ABA have been documented as evidence-based practices ([EBP] e.g. time delay procedures, prompting; Ferguson et al. 2020).

Being time- and cost-efficient, telehealth enables individuals with ASD to access EBPs more easily (Saral and Olcay 2021, Tomlinson et al. 2018). Recent reviews on telehealth-delivered interventions for individuals with ASD have revealed that telehealth appears to be a viable option for individuals with ASD to access a wider range of EBPs. Furthermore, these review studies indicate that the majority of the research focused on training clinicians or parents in implementing EBP through telehealth (Akemoglu et al. 2020, Ellison et al. 2021, Ferguson et al. 2020). In this regard, only a handful of studies have directly delivered interventions to individuals diagnosed with a disability via telehealth. Some of these studies have investigated the effects of an EBP on teaching skills (Cihon et al. 2022, Ferguson et al. 2020, Ford et al. 2021, Pellegrino and Reed 2020), while a few have investigated the effects of cognitive-behavioral interventions for illnesses such as anxiety and childhood insomnia (Hepburn et al. 2016, McCrae et al. 2021). Studies that taught participants with disabilities via telehealth used a system of least prompts to teach household skills (Pellegrino and Reed 2020), a training package comprising rapport-building, a system of least prompts, contingent praise, and refraining from negative comments for daily living skills (Ford et al. 2021), Cool Versus Not Cool for social skills (Cihon et al. 2022), discrete-trial teaching for language skills (Ferguson et al. 2020), and natural environment teaching and discrete trial training for adaptive, social, and language skills (Nohelty et al. 2022). The results of these studies indicated positive outcomes for participants of different age groups in terms of acquisition, maintenance, generalization, and social validity. In fact, telehealth-delivered interventions are effective for young children, teenagers, and adults with disabilities. This suggests that young children with ASD in particular can access training and education without the presence of their parents or caregivers, and that telehealth can be a viable alternative to face-to-face training during situations such as a pandemic.

The aforementioned research has established the effectiveness of telehealth-delivered interventions for individuals with ASD, but instructional efficiency should also be considered in the implementation process (Collins 2007, Vladescu and Kodak 2013). One way to achieve greater instructional efficiency is to conduct instructional sessions in a small group or triad arrangement for individuals with ASD in school environments (Ledford et al. 2008, National Research Council 2001, Tekin-Iftar and Birkan 2010). A triad arrangement is defined as a pair of students taught by one instructor (Wolery et al. 1991). This type of arrangement may be advantageous, as it provides students with an opportunity for observational learning and the acquisition of novel responses by observing the behaviors of a model (Bandura 1977). For example, instructional efficiency increases when students acquire information directly taught to other students during the triad instruction (Tekin-Iftar and Olcay-Gul 2016). Furthermore, delivering instruction in a triad arrangement facilitates peer interaction, thus minimizing the drawbacks of distance learning such as a lack of social interaction (Mengi and Alpdoğan 2020). Thus, evaluating the effects of a variety of EBPs delivered via telehealth may be important to disseminate these practices and provide another method of autism training among practitioners.

The simultaneous prompting procedure is one of the EBPs used to teach various skills to individuals with ASD across their age span (Collins 2007). The simultaneous prompting procedure involves two types of trials: daily probe trials and training trials (Collins 2012). Training trials involve presentation of the task direction (e.g. ‘Show me the red card’), followed by the presentation of the controlling prompt (e.g. the teacher immediately pointing to the red card) to ensure a correct response (Schuster et al. 1992). As controlling prompts are presented, and thus the student is not provided with an independent response in each training trial, daily probe trials are conducted to evaluate acquisition. Daily probe trials are conducted prior to each training trials to assess acquisition from the previous training session. Teaching sessions continue until the mastery criterion is met in the daily probe sessions (Tekin-Iftar et al. 2019). The simultaneous prompting procedure has several advantages compared to other response-prompting procedures. First, the simultaneous prompting procedure results in nearly errorless learning compared with other response-prompting procedures (Collins 2012). Furthermore, the use of a single controlling prompt makes the procedure faster-paced, as it does not require a prompt hierarchy, which results in significantly less time (Ault et al. 1988, Collins et al. 2018). Third, learners with ASD do not need the prerequisite skill of waiting for a prompt in simultaneous prompting. Finally, the procedure can be implemented with high accuracy in 1:1 and group instructional arrangements (Singleton et al. 1995). These characteristics, in combination with the effectiveness and efficiency of the procedure, make simultaneous prompting advantageous procedure (Waugh et al. 2011). Tekin-Iftar et al. (2019) conducted a meta-analysis of 20 simultaneous prompting studies employing a single-case methodology. They concluded that there is a rich body of evidence to support the procedure to teach individuals with ASD a variety of skills, including academic, daily living, and vocational skills, as well as to increase vocal responses. Among the studies reviewed, those aimed at increasing vocal responses taught expressive object naming (MacFarland-Smith et al. 1993), object identification (Tekin-Iftar et al. 2003), identification of mathematical symbols (Gursel et al. 2006), naming tools (Tekin-Iftar 2008), and telling the function of given organs (Tekin-Iftar and Olcay-Gul 2016) to participants with disabilities. The results of these studies indicate that the simultaneous prompting procedure delivered in person was effective in increasing the vocal responses of the participants. Moreover, a recent study revealed that the procedure was effective in increasing vocal responses (e.g. telling the characteristics of a good person) in four students with intellectual disabilities (Fidan and Tekin-Iftar 2022). These studies investigated the effects of a simultaneous prompting procedure delivered in person. Thus, no previous research has examined the effects of simultaneous prompting procedures delivered via in-person or telehealth in increasing the vocal responses of participants when playing sound. Furthermore, these studies examined the effects of the procedure on primary, secondary, and high school students with disabilities during 1:1 instruction. In this regard, to our knowledge, no studies have delivered simultaneous prompting directly to individuals diagnosed with a disability, particularly with ASD, during group instruction via telehealth.

Taken together, telehealth research has reported that (a) potential benefits and positive results of previous studies have led to widespread use of telehealth, (b) as only studies that delivered interventions directly to individuals with disabilities, additional research is needed; (c) there is a need for future research investigating the effects of different procedures from those in previous research directly delivered to young children with ASD, and (d) there is a need for research focusing on maximizing peer interaction during telehealth. Furthermore, the literature on simultaneous prompting shows that studies need to investigate the effects of telehealth-delivered simultaneous prompting procedures for young children with ASD during triad instruction. This study was conducted to fill the gap in the literature concerning the aforementioned rationales. Therefore, the following questions guided this study:

• Research Question 1: Is simultaneous prompting directly delivered to young children with ASD via telehealth during triad instruction effective in terms of the acquisition, maintenance, and generalization of the target information?

• Research Question 2: Is simultaneous prompting directly delivered to young children with ASD via telehealth during triad instruction effective in terms of the acquisition, maintenance, and generalization of the target information of the other student?

• Research Question 3: What are the opinions of the participating children’s teachers about the study in terms of (a) the effectiveness of the procedure, (b) the acceptability of the procedure, and (c) the social significance of the target skills for the students?

Method

Participants

An education coordinator in a foundation school, the second author, contacted the researchers and stated that she required directions to teach her students remotely because of frequent lockdowns due to the COVID-19 pandemic. Thus, the researchers conducted this study to provide her with the opportunity for effective and efficient instruction. The researchers first obtained institutional ethical approval, and then the education coordinator provided her students’ families with verbal and written information regarding the aims, significance, and procedures of the study. Two families provided written informed consent for their children to participate in the study. Moreover, verbal information about the study was provided to the participating students who assented to participate. Finally, predetermined prerequisite skills were evaluated to determine whether students were eligible for the study.

The following prerequisite skills were identified for the students: being able to (a) imitate 3-word verbal expressions, (b) follow simple directions, and (c) engage in an activity held long-distance. The prerequisite skills were determined based on the characteristics of the study’s dependent and independent variables. In other words, the target skill was to emit vocal responses, and the controlling prompts included verbal prompts; thus, the first prerequisite skill was imitating verbal expressions. Additionally, the implementer presented the participants’ skill direction to emit the correct responses, and the independent variable was delivered via telehealth. Therefore, the participants were required to have the second and third prerequisite skills. In evaluating the verbal imitation skills, the implementer emitted the student a total of five verbal expressions and asked him to repeat after her (e.g. ‘Repeat after me. “Cat says “Meow.”’’) To evaluate the ability to follow simple directions, the implementer verbally presented the student a total of five skill directions (e.g. ‘Place the book on the table’, ‘Close the classroom door’). The criterion was five correct responses out of five trials per session for both prerequisite skills. The implementer read a story of 20 sentences to the student via the Zoom Video Communications, Inc. (Zoom; www.zoom.us) videoconferencing platform to evaluate their ability to engage in an activity. The criterion was attending to the activity by looking at the screen and listening to the implementer until she finished the story in one session. Furthermore, the implementer asked the students five questions related to the story. The criterion was to answer all questions correctly.

Two boys aged 5 diagnosed with ASD participated in the study (see Table 1). Neither student had a prior history of learning through telehealth, but they had a history of educational intervention that used a prompting system. Ali was a 5-year-old boy who had been diagnosed with ASD at the age of 28 months. He attended a school for individuals with ASD affiliated with an autism foundation on a full-time basis, and attended a private preschool inclusive classroom on weekends. His score on the Gilliam Autism Rating Scale-2, Turkish Version (GARS-2-TV), indicated a high likelihood of autism. In addition, his score on the Test of Early Language Development-Turkish Version (TELD-TR; Topbas and Guven 2011) indicated a receptive and expressive language age of 38 months. He could follow simple directions, respond to questions, and make requests but experienced difficulty in following classroom rules, taking turns, and initiating interactions. The other student, Mert, was a 5-year-old boy who had received an ASD diagnosis at the age of 30 months. He attended a school for individuals with ASD affiliated with an autism foundation on a part-time basis, and attended a private preschool inclusive classroom on weekends. His GARS-2-TV score indicated a very high likelihood of autism. His TELD-TR score indicated a receptive language age of 49 months and expressive language of 38 months. He could follow simple directions, respond to questions, and initiate and maintain interaction, but had difficulty following classroom rules, having conversations appropriate to the context, and exhibiting adaptive social skills. Both students were able to match objects and their pictures, show a different one, match the same sounds, and discriminate between the different sounds. However, they had difficulties in grouping objects and their pictures based on various characteristics and indicated the being (e.g. animal) and the category based on a sound.

Table 1.

Descriptions of child participants.

Child	Diagnosis	Gender	Age	TELD-TR		GARS-2-TV
				RLD	ELD
Mert	ASD	Male	60 months	Under the average point in RLD. Equivalent to 49 months	Under the average point in ELD. Equivalent to 38 months	71 – Very likely probability of autism
Ali	ASD	Male	60 months	Under the average point in RLD. Equivalent to 38 months	Under the average point in ELD. Equivalent to 38 months	88 – Very likely probability of autism

Note. ELD: expressive language development, RLD: receptive language development.

Implementer and observer

The second researcher, who was the education coordinator in the foundation school, implemented all experimental sessions remotely. She was also the primary data collector for this study. She was a master’s student in special education and had 11 years of experience working with individuals with ASD using response-prompting procedures. An observer who was another teacher in the same school collected the dependent and independent reliability data. The researchers explained to the observer how to collect reliability data regarding the following subjects: (a) the goals of the study, (b) target behaviors, (c) simultaneous prompting delivered via telehealth during triad instruction, (d) experimental sessions, (e) response definitions, and (d) implementer behaviors.

Setting

All sessions were conducted using the Zoom videoconferencing platform. Students participated in the sessions using a laptop computer. The students and implementer were located at different locations. Both students were located within a separate classroom in the school, as all lessons were held individually due to COVID-19. The implementer was located at home. Teachers were in the same classroom in the case of a technical problem. However, no such problems were observed.

Materials

PowerPoint slide presentations were prepared for all experimental sessions. The presentations included visual and audio stimuli. To list three appropriate exemplars for a given category, the presentation slides included one visual for the corresponding question (e.g. a key picture for ‘Tell three things to open with a key’) during the probe, maintenance, and generalization sessions. The presentation slides included three additional stimuli (i.e. a house door, padlock, and car door) for the corresponding question during the instructional sessions. To respond to animal/object sounds, the presentation slides included only an audio icon for the corresponding stimulus during probe, maintenance, and generalization sessions, and included one visual of the animal/object (e.g. a motorcycle picture) during instructional sessions. An example presentation slide used during the instructional sessions is shown in Figure 1. The presentation slides included three sets of three target stimuli for a total of nine slides.

Figure 1.

An example of a presentation slide for the ınstructional sessions.

Experimental design

A multiple probe design across behaviors, replicated across participants, was used to examine the effects of simultaneous prompting delivered young children with ASD via telehealth (Ledford et al. 2019).

Dependent variables

The dependent variables were selected considering the monthly targets according to the students’ Individualized Education Program (IEP). In fact, these targets were planned to be taught to students within that month based on their IEPs. As additional data were collected on student acquisition of observational responses, the researchers chose the target stimuli for areas of overlap across students. The dependent variable was correct and unprompted vocal responses by students (listing three appropriate exemplars for the given category for Mert and responding to animal/object sounds for Ali; Table 2). A correct response was defined as giving a correct example related to the question for Mert and responding to animal/object sounds for Ali within 4 s. An incorrect response was defined as not giving a correct or incorrect example related to the question for Mert and not responding or responding differently to the animal/object sound for Ali within 4 s.

Table 2.

Target ınformation for the students.

Mert			Ali
Listing three appropriate exemplars for the given category			Defining the name of being sounds
Set 1	Set 2	Set 3	Set 1	Set 2	Set 3
1. Tell three things to sit on.	4. Tell three things to open with a key.	7. Tell three things to wear on feet.	1. Door sound.	4. Plane sound.	7. Train sound.
2. Tell three things that fly.	5. Tell three things with a tire.	8. Tell three things with a moustache.	2. Grasshop per sound.	5. Drill sound.	8. Owl sound.
3. Tell three vegetable names.	6. Tell three domesticated animals.	9. Tell three things that grow on ground.	3. Motorcycle sound.	6. Wolf sound.	9. Hairdryer sound.

Independent variable

The independent variable was the simultaneous prompting procedure and reinforcement delivered via telehealth during triad instruction via the Zoom videoconferencing platform.

General procedure

The screening, probe, maintenance, and generalization sessions were conducted during 1:1 instruction. Instructional sessions were conducted during the triad instructional arrangement. All correct behaviors resulted in praise in all sessions except the screening sessions. The implementer ignored incorrect responses and no responses in any of the sessions.

Screening sessions

Prior to the baseline sessions, the implementer conducted two consecutive screening sessions to identify the target stimuli for each student. A total of 12 stimuli were selected based on the school’s curriculum. Each stimulus was included in the PowerPoint slides and used to screen for unknown stimuli by having Ali and Mert receptively identify known stimuli. Three stimuli from a pool of unknown stimuli were selected as target stimuli for each student. Three trials were conducted for each screening session. First, the implementer secured the students’ attention (that is, ‘I will ask you some questions. Ready to work with me?’). When the student expressed their readiness with a signal or verbally, the implementer provided verbal praise (e.g. ‘Great!’). The implementer then shared the screen in Zoom and started the presentation. The implementer presents the skill direction (e.g. ‘Tell three things to wear on the feet’) and waited for the student’s response for 4 s before starting the next trial.

Probe sessions

There were two types of probe sessions: full and daily. The implementer behaviors during these sessions were as follows: (a) delivering the attentional cue, (b) reinforcing students’ affirmative response, (c) sharing the screen, (d) presenting the skill direction, (e) waiting for a 4-second response interval, (f) delivering the appropriate consequence, (g) waiting for a 4-second intertrial interval, and (h) recording students’ responses. Three trials were conducted for each target stimulus during each session.

Full probe sessions

Full probe sessions were conducted before each intervention phase to determine current performance and after the participant met the mastery criterion to evaluate whether the correct response was maintained. There was one full probe session per day, and each probe session continued until a contra-therapeutic effect was evident for three consecutive sessions.

In a full probe session, the implementer secured the students’ attention (e.g. ‘I will ask you some questions. Ready to work with me?’). When the student expressed his readiness with a signal or verbally, the implementer provided verbal praise. The implementer then shared the screen in Zoom and started the presentation. The implementer presented the skill direction (e.g. ‘Listen to the sound. What sound is it?’) and waited for the student’s response for 4 s before starting the next trial. Correct responses resulted in verbal praise (e.g. ‘Yes! Great!’). The implementer ignored incorrect and no responses and started the next trial. A plus (+) was used to indicate correct responses, and a minus (−) for incorrect or no responses on the data collection form, which was not presented to the students.

Daily probe sessions

Simultaneous prompting does not allow the evaluation of participant performance on acquisition within an intervention session, because the skill direction is immediately followed by a controlling prompt. Thus, daily probe sessions, in which no prompts were presented, evaluated the acquisition level of the participants. The daily probe sessions were the same as the full probe sessions, with the only difference that student performance was evaluated for all target behaviors during full probe sessions, whereas student performance on the target behavior currently taught was evaluated during daily probe sessions. Each daily probe session was conducted immediately before the instructional sessions and consisted of nine trials, with three trials per target stimulus. Daily probe sessions were continued until the students exhibited correct responses with 100% accuracy for three consecutive sessions.

Observational learning probe sessions

During these sessions, to probe observationally learned responses, the implementer collected data on whether the student acquired the information that was targeted for the other student. These data were collected during observational learning probe sessions conducted immediately after the sessions. The procedure in the probe sessions was followed in observational learning probes. Each session consisted of nine trials with three trials per target stimulus.

Instructional sessions

This study examined the effects of simultaneous prompting and reinforcement on teaching target behaviors in young children with ASD. The implementer behaviors during instructional sessions were as follows: (a) delivering the attentional cue, (b) reinforcing students’ affirmative response, (c) sharing screen, (d) presenting the skill direction, (e) delivering the controlling prompt (visual and verbal prompts), (f) waiting for a 4-second response interval, (g) reinforcing correct responses, (h) ignoring incorrect and no responses, (i) waiting for a 4-second intertrial interval, and (j) recording students’ responses.

The implementer secured the students’ attention (i.e. ‘I will ask both of you some questions. Ready to work with me?’). When the students expressed their readiness with a signal or verbally, the implementer provided verbal praise. While working with one of the students, the implementer encouraged the other to observe the instruction (i.e. ‘Watch me and your friend very carefully’). The order of the stimuli was randomly determined prior to each session. The implementer shared a screen containing a PowerPoint presentation with nine slides (i.e. one slide per trial). Presenting the student a skill direction (e.g. ‘Listen to the sound. What sound is it?’ and playing a train sound), the implementer provided the controlling prompt with no delay. For this purpose, she presented visual and verbal prompts by clicking on the screen to allow the visuals to show (e.g. a train picture) and saying ‘a train’. Then, the implementer waited for the student’s response for 4 s. Correct responses resulted in verbal praise (e.g. ‘Yes! Great!’). The implementer ignored incorrect and no responses and went to the next slide, which included the stimuli for the other student. The order of students was randomly determined prior to each session.

Maintenance and generalization sessions

The researchers conducted an additional maintenance phase, although post-intervention probe phases serve as a maintenance condition for the tiers in which the criterion had been met. The maintenance sessions were conducted at one, two, and six weeks after the intervention phases ended.

The generalization sessions were remotely conducted before and after the intervention as pre- and post-tests across people and stimuli by a different teacher in the school, using stimuli of different sizes and colors from the instructional sets. A different teacher in the school served as an implementer and used a PowerPoint presentation that included larger and smaller stimuli with different colors in generalization sessions than in instructional sessions. Pre-test generalization sessions were conducted during the initial probe phase. Post-test generalization sessions were conducted once the mastery criterion was reached during the intervention. The maintenance and generalization sessions were the same as the probe sessions.

Social validity

Social validity was assessed by interviewing the students’ teachers. The researchers developed a social validity form that included five open-ended questions on the following topics: (a) effectiveness of the procedure, (b) acceptability of the procedure, and (c) social significance of the target skills for students. Table 3 includes the social validity questions.

Table 3.

Social validity questions.

1. Do you think the target skills were functional for the student participants?

2. What do you think about simultaneous prompting procedure delivered via telehealth?

3. Do you think telehealth is an effective and viable option to teach your students a variety of skills?

4. What did you like most about the training?

5. What did you like least about the training?

Data collection and analysis

Three types of data were collected and analyzed in the study: reliability data (interobserver agreement, procedural fidelity), effectiveness data, and social validity data.

Reliability analysis

All sessions were screen-recorded to assess inter-observer agreement (IOA) and procedural fidelity (PF) data. The observer independently recorded the students’ responses for 100% of the sessions per phase for each student. IOA percentages were calculated using a point-by-point agreement formula: dividing the number of agreements by the number of agreements and disagreements, and multiplying by 100 (Cooper et al. 2019). The mean IOA was 100% in all sessions. PF data were calculated using the following equation: the number of observed implementer behaviors divided by the number of preplanned implementer behaviors and multiplied by 100 (Cooper et al. 2019). The procedural fidelity was 100% for all sessions.

Effectiveness data analysis

The students’ target and observational responses were visually inspected to determine the potential effect between the telehealth-delivered simultaneous prompting and reinforcement during the triad instruction and the students’ responses. The researchers evaluated the level, trend, variability, immediacy of the effect, overlap, and consistency of data patterns for the visual analysis of the data. as suggested by Kratochwill et al. (2013). Next, the effect size scores of each student’s target responses were calculated using the singlecaseresearch.org Tau calculator (Vannest et al. 2016) to quantify the magnitude of change that occurred between baseline and intervention. Tau-U scores can be interpreted as follows: weak or small effect: ≤ 0.65, medium to high effect: 0.66–0.92, and large or strong effect: 0.93–1.00 (Rakap, 2015).

Social validity data analysis

Social validity data collected from the participants through semi-structured interviews were descriptively analyzed.

Results

The percentages of correct responses for Ali and Mert are shown in Figures 2 and 4, respectively. Figures 3 and 5 show the percentages of correct observational responses for Ali and Mert, respectively. Both participants responded incorrectly during all pre-intervention phases, reached the mastery criterion during the intervention, generalized the acquired behaviors, and maintained them when the intervention was removed.

Figure 2.

The percentages of correct responses for Ali.

Figure 4.

The percentages of correct responses for Mert.

Figure 3.

The percentages of correct observational responses for Ali.

Figure 5.

The percentages of correct observational responses for Mert.

During the initial probe phase, Ali did not respond to the animal/object sounds. He only responded with 33% accuracy in one session during the second and third probe phases. During the intervention for Set-1, there was an immediate upward trend in his correct responses, with no overlap with the baseline data. During the intervention for Set-2, his correct responses abruptly increased, followed by variability for four sessions. His correct responses increased in level and variability with the introduction of the intervention for Set-3. Ali reached the mastery criterion for responding to animal/object sounds in four sessions for Set-1 and eight sessions each for Set-2 and Set-3.

Regarding listing three appropriate exemplars for a given category acquired through observational learning, Ali did not list any exemplars during the pre-intervention probe sessions. With the introduction of Set-1, his correct responses increased, which then fell by half. There was an immediate upward trend in correct responses for the subsequent sessions. During the intervention for Set-2, Ali did not exhibit any correct responses in the two sessions. Subsequently, his correct responses increased in level following a gradual increase. For Set-3, the correct responses showed increased variability over several sessions. An increasing trend in his correct responses was evident in subsequent sessions.

Mert did not list any exemplars for any given category during the initial probe phase. He only responded with 33% accuracy in one session during the third probe session. During the intervention for Set-1, an increasing trend in his correct responses was evident. With the introduction of the intervention for Set-2, his correct responses showed some variability over several sessions and continued to increase. This pattern was replicated in Set-3. Mert reached the mastery criterion for listing three appropriate exemplars for the given category in five sessions for Set-1, and six sessions for Set-2 and Set-3, each.

Regarding the response to animal/object sounds acquired through observational learning, Mert did not respond to animal/object sounds during the pre-intervention probe sessions for Set-1. However, he responded with 33% accuracy during the second probe sessions for Set-2 and third probe sessions for Set-3. With the introduction of Set-1, his number of correct responses gradually increased. With the introduction of the intervention for Set-2, his correct responses showed some variability over several sessions and continued to increase. This pattern was replicated for Set-3.

Regarding the effect size, Mert demonstrated a medium to high effect size across sets. The effect size was 0.86 for Set-1, 0.88 for Set-2, and 0.85 for Set-3. The mean effects demonstrated a medium to high effect size of 0.86. Furthermore, Ali demonstrated a large or strong effect for Set-1 (1.00) and Set-2 (0.98). For Set-3, the effect size was 0.88. The mean effects demonstrated a large or strong effect of 0.95 (see Table 4).

Table 4.

Tau-U effect size on target behaviors in students.

Student	Set 1	Set 2	Set 3	Mean
Mert	.86*	.88*	.85*	.86
Ali	1.00*	.98*	.88*	.95

Note. ^*p < 0.001.

For Ali, maintenance data indicated that he maintained the target behaviors across all sets, responding to animal/object sounds, at the mastery criteria. Similarly, the target behaviors for Mert were maintained at the mastery criterion. Both students maintained behaviors acquired through observational learning with 100% accuracy. During the pre-test generalization sessions, neither student exhibited the correct behaviors. During the post-test generalization sessions, the data showed that both students generalized the acquired target behaviors with 100% accuracy across different stimuli and teachers. They also generalized the behaviors acquired through observational learning across different teachers.

Social validity

The teachers indicated that the study contributed to their students’ overall performance. Furthermore, they stated that the target behaviors were functional; as such, their students would be able to use them in daily life. Regarding the negative aspects of telehealth, teachers reported that distance education might not be effective for young students who cannot maintain their attention for longer periods. For example, Ali’s teacher noted that the instructional sessions may be too short if the students cannot sufficiently focus on the instruction on the screen. The teachers agreed on the positive aspects of telehealth, indicating that students who cannot participate in classroom activities for any reason, such as a pandemic or illness, may have the opportunity for peer interaction via telehealth. Finally, the teachers stated that telehealth was a viable option as an educational support model for teaching a variety of skills to young children with ASD. For example, Mert’s teacher said, ‘In today’s world where smartphones and iPads are a must for everyone, distance education is inevitable. Therefore, telehealth is an effective practice for access to training for any student’.

Discussion

This study aimed to examine the effects of telehealth-delivered simultaneous prompting and reinforcement in teaching target behaviors in young children with ASD. Target and observational responses were acquired by the students, maintained, and generalized across different conditions. The effect size results for the students indicated large or strong effects and gains, as indicated by visual inspection of the data.

Previous research on telehealth in teaching children with ASD has largely focused on decreasing or eliminating problem behaviors (e.g. Lerman et al. 2020, Wacker et al. 2013). In fact, only a handful of studies utilized telehealth to document validated practices for skill acquisition in children diagnosed with ASD (e.g. Sivaraman et al. 2021). This shows the need for additional research, since telehealth may function as a feasible service delivery option to implement EBPs.

The current study supports previous research showing the effectiveness of simultaneous prompting in teaching young children with ASD. The effectiveness of the simultaneous prompting procedure has been extensively documented (Tekin-Iftar et al. 2019). However, to our knowledge, this is one of the few studies to implement EBP through telehealth for individuals with ASD (Cihon et al. 2022, Ferguson et al. 2020, Ford et al. 2021). Furthermore, one unique characteristic of the current study is the use of telehealth to directly implement a simultaneous prompting procedure for this population. This suggests that in-person simultaneous prompting for young children with ASD can be achieved via telehealth. Given the limited research on interventions that are directly delivered to young children with ASD via telehealth, this study advances the literature on telehealth and provides interventionists with a feasible alternative option for in-person interventions.

The data suggest that both students were able to acquire observational responses through observational learning. In other words, each student acquired the information that was taught to the other student in the triad at the criterion level. There were no correct observational responses during the pre-intervention probe sessions for Ali and 33% accuracy during two probe phases for Mert. However, observing the behaviors taught to the other student in the triad resulted in 100% accuracy for both students. The intervention sessions were very short, and both students required less than eight sessions to acquire the nine responses, suggesting that the intervention was both effective and efficient. These outcomes are similar to those in previous studies on simultaneous prompting delivered in-person (e.g. Gursel et al. 2006, Palmer et al. 1999). The findings also extend the literature on applied behavior analysis telehealth, given that only one such study delivered a procedure directly to young children with ASD via telehealth and measured observational learning of stimuli assigned to peers (Ferguson et al. 2020). Ferguson et al. (2020) investigated the effects of discrete-trial teaching with instructive feedback delivered via telehealth to teach tact relations to dyads of children with ASD. They found that five of the six participants acquired observational responses. Clearly, there is a need for additional research to further investigate the effectiveness and efficiency of interventions directly delivered to individuals with ASD to inform future efforts focused on promoting an evidence base of telehealth.

The effectiveness of the simultaneous-prompting procedure has been demonstrated in one-to-one settings (e.g. Tekin-Iftar et al. 2017), small-group settings (e.g. Gursel et al. 2006), and large-group settings (e.g. Johnson et al. 1996). To our knowledge, only one study has examined the effects of an intervention delivered via telehealth to teach individuals with ASD during triad instruction (Ferguson et al. 2020). The researchers taught tact relations to six children diagnosed with ASD using discrete trial teaching, with instructive feedback delivered directly via telehealth. They reported that all participants acquired their primary and secondary targets. Therefore, one unique aspect of the current study is the use of the telehealth setting to directly implement a procedure for a triad with young children with ASD. Considering the recent increased need for effective interventions delivered via telehealth, the findings provide evidence that the telehealth format was effective for a triad with participating young children with ASD. However, future research is needed to compare the interventions delivered via telehealth and in person to provide empirical evidence for the use of one procedure over the other during teaching triad instruction.

Both students maintained their target behaviors and observational responses after the intervention. All studies evaluating the use of telehealth to directly teach individuals with a disability measured maintenance of the target skills (Cihon et al. 2021, Ford et al. 2021, Pellegrino and Reed, 2020) and observational responses (Ferguson et al. 2020), and reported positive outcomes for all participants. However, only one study measured long-term maintenance (6–11 weeks) for two participants (Ford et al. 2021), which reveals a need for future studies that measure the long-term maintenance effects of treatment gains.

The researchers also assessed the generalization of the target skills, in addition to the behaviors acquired through observational learning. Telehealth-delivered simultaneous prompting procedures and reinforcement were effective for the students’ generalization of the target skills and behaviors acquired through observational learning. Cihon et al. (2021) reported that two of the three participants generalized target skills across adults. In another study, participants generalized the target skills across materials (Ford et al. 2021). However, this literature base can best be described as limited, owing to the limited number of studies and participants. Future research should continue to measure the generalization of target skills and observational responses. Additional research is needed to measure generalization outside of the telehealth direct setting.

The social validity results showed that the intervention was effective and acceptable, and the target skills were significant for students. Furthermore, the intervention was an acceptable replacement for in-person instruction for the respondents. These results support the idea that interventionists should deliver simultaneous prompting procedures via telehealth. However, these results should be interpreted with caution because the data were collected from a limited number of respondents in this study. Future research should measure the social validity of the intervention by examining the opinions of participants themselves and other family members, such as siblings and parents of participating students, who are less likely to be biased.

As anecdotal data show, Mert always verbally interacted with Ali during the sessions. At the beginning of each session, he greeted Ali and spoke to him. His mother reported that he looked forward to the start of sessions. Moreover, Mert always followed the implementer’s instructions during all the sessions. However, Ali never initiated an interaction or responded to his peer’s interactions during the first several instructional sessions. Later, he started to initiate verbal interaction by emitting expressions such as ‘Hello’, ‘Goodbye’, or ‘How are you?’ modeled by his peer. Furthermore, his teachers reported that he asked about Mert after the conclusion of the intervention. Both students showed high attendance and compliance rates. More systematic data collection should be performed to evaluate participants’ attendance, compliance, engagement, or challenging behaviors during a telehealth-delivered intervention, which may be significant for students to learn through direct telehealth intervention.

Caution is needed regarding the suitability of direct telehealth services for individuals with ASD. The Council of Autism Service Providers (2020) suggested prerequisite skills for participation in a telehealth direct service, such as attending to and following directions and sitting in front of the computer. In this study, we determined the prerequisite skills by considering the characteristics of the dependent and independent variables. The participants were required to imitate verbal expressions, follow simple directions, and sit in front of a computer to exhibit the target responses. Thus, the findings of this study may be generalized across individuals with ASD having these prerequisite skills, yet it warrants further investigation and future replication studies using the same prerequisite skills to support this.

These findings add to the growing body of literature on telehealth among young children with ASD. However, this study has several limitations. First, the students’ teachers arranged the Zoom calls for them. In other words, an additional person is required in the classroom to provide technical support. If telehealth is a viable practice for teaching young children with ASD, the effectiveness of the interventions should be investigated in teaching them how to use telehealth. Second, only two young children with ASD participated in this study. Therefore, these findings are limited by their characteristics. Third, as the third researcher collected data on social validity;, the data collection was not anonymous. Future studies should collect social validity data through measures that are less subject to bias, such as blind ratings, which may provide more valid evidence of what components (i.e. goals) are socially acceptable and significant.

Despite these limitations, these results may be generalized to other young children with ASD who have no history of telehealth-delivered intervention, since previous studies investigated the effects of telehealth-delivered interventions for teaching individuals diagnosed with a disability with experience with interventions delivered via telehealth (Cihon et al. 2021, Ferguson et al. 2020, Ford et al. 2021, Pellegrino and Reed 2020). Telehealth may be a promising practice enabling professionals to teach young children diagnosed with ASD, including those who may otherwise be unable to access effective instruction due to distance. However, further research is needed to investigate the effects of interventions delivered via telehealth due to the limited literature.

Disclosure statement

The authors have no relevant financial or non-financial interests to disclose.